Recently high-carbohydrate, low-fat diets have been widely advocated for diabetic patients (3,17,20,23,24,30,36,48,52). As arguments for this therapy it has been claimed that hypercholesterolemia is of grave prognosis in diabetes and that high fat diets may superinduce hyperlipemia. The following investigation was undertaken to determine the incidence of hypercholesterolemia in diabetes, to study the relation of the cholesterol level to that of fatty acids and phospholipoids in blood serum, and also to learn whether diets moderately high in fat such as are used in this clinic give rise to hyperlipemia.
MATERIALS AND METHODSThe serum, from venous blood taken before the morning insulin or breakfast, of 79 different diabetic patients, 20 males and 59 females of all ages, has been examined 130 times for cholesterol, lipoid phosphorus and non-phospholipoid fatty acids by methods already described (37,38). The percentage errors of the analytical techniques have been discussed previously (37,38,40). In almost every instance serum proteins (11) and blood sugar (4, 54) were determined simultaneously. No studies which were made when the patients were dehydrated, acidotic, pregnant, postpartum, or had been treated with intravenous injections of acacia are included. DATA Data are given in Table I. Insulin dosage and protein, fat and carbohydrate intake indicate the requirements and food consumption of the patient for at least four or five days previous to the blood study. Whenever there was any doubt that the diet had been followed adequately, the term " unregulated " has been used. The blood of many of these patients was examined the morning after admission to the hospital, and unless there was an exact record of the diet the regimen was condemned to the "unregulated" category. Brief protocols concerning each patient are presented at the end of the paper. Notes are included where there were evidences of hypertension or arteriosclerosis, the latter being judged by evidences of sclerosis in the eyegrounds and extremities, cardiac enlargement and decompensation. Blood pressures were observed on all patients, not always on the day of the blood study, but when the subject was neither unduly excited nor in circulatory embarrassment. In addition to the actual body weight, the impression of nutritional state is described. Any evidences of liver pathology or enlargement, or of kidney pathology, are included, but neither organ is mentioned unless an abnormality was discovered. The albumin and globulin in the serum of more than half the patients were determined. In most instances any alteration in total protein was due to a variation in albumin, but in certain cases when the albumin: globulin ratio was abnormal the concentrations of the fractions are included.Cases have been grouped according to the concentration of cholesterol in the serum. Whenever this fluctuated the study used for classification was selected when the condition of the patient was most nearly normal or when the cholesterolemia had attained a constant level. The tab...